Tank Purging Calculations 1

[niallmacdubhghaill]

Hi,

Im working on trying to figure out how much N2 is required to inert a vessel such that the final O2 conc. within the vessel is <= 2%.

I have done some sums which indicate that ideally 3.5 tank volumes would be required to do this. The other engineers in my office generally agree that this is about right. (Anyone with any comments otherwise, please let me know)

The physical set up in the tanks for this purge is such that the N2 inlet and outlet are at the top of the tank.

My concern is that this arrangement would result in the possibility of a "short circuit" occuring, and that instead of achieving proper mixing and thus a uniform conc of O2 throughout the tank, there may be areas in the tank in which the O2 conc is above the 2% requirement.

There are many different tanks with many different geometries in this plant, torispherical, conical, flat bottoms and tops.

I was wondering if anyone knew of any correlations that would provide a "mixing efficiency" that would tell me what to multiply by to ensure my answer is as close to correct as possible. That is, something that would take into account the required final atmospheric composition, the vessel geometry, filling/purging rates.

For the record, the following options are not viable for operational reasons:

We cannot fill the tank with liquid first and then use N2 to displace the fluid

We cannot introduce the N2 at the bottom of the tank.

The option of running the agitators during the inertion/purging operation to promote turbulance/mixing. Does anyone have any comments on this idea, i.e. is it a good/bad/indifferent idea to run an agitator like this when it is designed to agitate a liquid at least as viscous as water, sometimes much more so.

Please note that this is a very general question as we are dealing with many different vessels & configurations.

Any input is greatly appreciated.

 

[25362]

[&bull;] If the vessel could stand some pressure, the pressure purging technique would probably ensure complete mixing.

The number of cycles would then be computed from:

[ln (2[&divide;]21)] [&divide;] [ln (1.0[&divide;]P)]​

Where 1.0 is the starting pressure (ie, atmospheric), P is the absolute pressure to which the vessel is brought on each cycle. If, for example, P can be 2 ata., the number of cycles would be:

-2.35 [&divide;] -0.693 = 3.4, namely 4 cycles, to reduce oxygen from 21 to 2%.

The higher the pressure that the vessel can sustain the less the number of pressure cycles.

[&bull;] Depending on the size of the vessel and whether its design includes full vacuum, vacuum purging could be your solution.

 

[niallmacdubhghaill]

25362,
Unfortunately, the tanks are being designed to withstand only a very small amount of over pressure, once the pressure gets above 1.03BarA there is a relief valve that goes.
What I was wondering was if you knew of any fudge factor that I could use to allow for incomplete mixing etc?
Cheers
Niall

 

[25362]

Here are some comments for your consideration:

1. Can you introduce nitrogen through a flexible hose through the inlet nozzle ?

2. Since the principle of "safety first" governs, the best you could do is by trial.

3. By putting in 3.5 tank volumes you are already using a "safety" factor of ~1.5 in your sweep-through purging scheme.

The theoretical volume, Q, of pure nitrogen assuming perfect mixing inside the tank of volume V, would be:

Q = V ln(21[&divide;]2) = 2.35[&times;]V​

4. As for the short-circuiting effect consider the following aspects:

[&bull;] The average speed of nitrogen molecules in air at 25^oC is 515 m/s. With velocity being proportional to the square root of the absolute temperature.

[&bull;] The average speed of molecules in a gas is inversely proportional to the square root of the molar mass. Thus

v(N₂)/v(air) = [&radic;] (30[&divide;]28) = 1.035​

meaning that nitrogen molecules move 3.5% quicker than air.

[&bull;] Nitrogen and moist-saturated air have about the same density, and if nitrogen is a bit cooler its density may even surpass that of dry air.

Good luck.